Among all the damages to the genome, DSBs are considered one of the most deleterious to cells. They arise from both environmental agents like ionizing radiation or chemotherapeutic drugs and normal cellular processes like DNA replication, V(D)J recombination, or meiosis. If un-repaired or improperly repaired, DSBs would cause chromosome deletions or translocations, ultimately leading to premature cell death or oncogenic transformation. Mutations in many DSB repair genes, such as Werner syndrome gene and Bloom syndrome gene, BRCA1, and BRCA2, dramatically increase the risk of cancer. Clinically, eliciting DSBs by ionizing radiation and various cancer drugs is also among the most commonly used methods to treat cancer. DSBs are repaired by two general types of pathways: non-homologous end joining (NHEJ) and homology-dependent repair (HDR). Proper choice of repair pathway is critical to genome stability. The key event in the bifurcation of the two pathways is the initial processing of DNA ends. NHEJ involves limited processing, but HDR requires extensive processing to form 3' ss-tails. Recent studies from several labs, including mine, have elucidated the basic mechanisms for resection. However, many important questions are still poorly understood. Firstly, HDR and NHEJ are both active during S-G2 phases of the cell cycle, but it is unclear what factors first determine if a DSB is channeled to resection (for HDR) or to NHEJ. Secondly, ends generated by ionizing radiation and many cancer drugs often carry damaged nucleotides. These ends can still be stably bound by NHEJ factors, but repair cannot be completed or is seriously delayed. It is unclear if they are trapped or can be re-channeled to resection for HDR. Thirdly, compared to model systems, the understanding of resection in human cells is still very limited. In this application, three specific aims are proposed to address these important questions. Specific Aim I is designed to test the hypothesis that a key factor for determining if a DSB is resected or not is the structure of ends. DNA with ends linked to a protein adduct, which are frequently induced by many cancer drugs, will be used as a model substrate to test this hypothesis. Its repair will be rigorously analyzed by biochemical reconstitution studies in Xenopus egg extracts and with purified resection proteins. Specific Aim II is designed to test the hypothesis that ends with damaged nucleotides are bound by NHEJ factors but then re- channeled to resection for HDR by the MRE11-RAD50-NBS1 (MRN) complex. The target protein that is dislodged from ends by the MRN-mediated mechanism will be identified. Specific Aim III is designed to test the hypothesis that the Werner syndrome protein (WRN) and the DNA2 nuclease, which are critical for resection in Xenopus egg extracts, are also important for resection in human cells. These studies will greatly increase the understanding of how DNA ends are resected and consequently how DSB repair pathways are chosen. The key proteins involved in DSB resection and end re-channeling might be developed into new targets for drugs or as biomarkers to increase the efficiency of radiation therapy and chemotherapy of cancer.